A recent paper from our lab - now also a thread on @Mastodon 🚨
The most important readout in any preclinical #therapy for #stroke⚡️is beahvior. Still, tests in 🐭are poorly reproducible + often unsuitable for unraveling complex recovery.
We tried to change that using #DeepLearning, now out in
#BMCBiology
https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-022-01434-9 🧵
@ruslan
These are really cool results!! I wonder if you tried Anipose for reconstructing the 3D pose and joint angles of the mouse?
@neurolili thank you! I know many people use Anipose and are also happy with the results - so I assume you may get comparable results. However, we did not do a direct comparison of the methods.